Existing computing architectures for managed-account transactions present many shortcomings. For example, conventional architectures are typically based exclusively on settlement-date data and utilize shadow accounting, which often presents reporting and reconciliation challenges, both in the performance and regulatory contexts. These challenges are exacerbated by the failure of such architectures to utilize data schemas and structures designed from the outset for managed-account transactions, as well as a lack of appropriate data quality checks throughout their lifecycles. Accordingly, a need exists for an improved computing architecture for managed-account transactions.